Migration settlement networks in the Carpathian Basin, 2001–2011 Áron Kincses Looking at the relationship between the place of Hungarian Central birth and current residential locations of foreign Statistical Office, citizens arriving in Hungary from the neighbouring Budapest, Hungary countries, in general, we establish that smaller E-mail: migration distance involves migrants with a lower [email protected] level of education, while preference for longer distances is determined by higher qualifications of migrants. The potential impact area of migrants Lajos Bálint grows in line with the education attainments of Hungarian Demographic migrants. Research Institute A scale-free settlement topology can be seen Budapest, Hungary from the neighbouring countries of immigration to E-mail: Hungary. This means that most of the settlements [email protected] of Hungary have just a few links to settlements of neighbouring countries, from a migration point of view, while few Hungarian settlements have many connections. This finding also means that, instead of Keywords: the national migration strategy, the subsidiary and international migration, the regional strategies can play a decisive role in the Carpathian Basin, management of the international migration process. network analysis Introduction The key to migration settlement networks is linking the source countries to the destination areas. The analysis is based on the data on foreign nationals that is collected from the census and the Office of Immigration and Nationality. Establishing the linkage between the two databases has created an opportunity to connect the study on migration settlements with the detailed information material of the census on the subject. In this study, we strive to show in detail the characteristics of the international migrants’ settlement network in relation to Hungary and its neighbouring countries and connect them to the attributes of migrants. Analysing the relationship between the Carpathian Basin migration source and destination areas, our aim is not only to explore the regional features of the flows Regional Statistics, Vol 6, No 2. 2016: 95–113; DOI: 10.15196/RS06205 96 Áron Kincses – Lajos Bálint between a given emigration country and Hungary but also to depict a general territorial network of contacts in an integrated way, simultaneously taking into account all the neighbouring countries. In this network, which is considered as the specific sub-network of the entire Carpathian Basin network, we will examine the settlement networks of the migrants, mostly of Hungarian ethnicity, arriving to Hungary. The question is whether it is characteristic that some parts of the aforementioned territorial network are increasingly ‘utilised’ by the foreign citizens living in Hungary and having specific demographic, labour market properties, or given territorial connections are independent of the characteristics of migrants. Data, relations of source, and destination areas Census data represent a detailed demographic, labour market, and sociological data set on the population of migrant origin living in Hungary. However, information on the places of emigration and birth only country is available. Relevant Hungarian migration databases (database of Office of Immigration and Nationality and the HCSO data files) contain less information on the characteristics of migrants, but extend to their places of birth. The establishment of link between the two databases enables to connect the analysis on the emigration areas with the detailed census information material. The method is based on the use of a complex conversion key between databases, which assigns the municipalities in foreign places of birth1 to the census data files according to the common variables (nationality, residence in Hungary, date of birth, gender, and marital status). Thus, although data on foreign settlements underlying the specific analyses was available, a separate classification became necessary. It is because such classification often contained the denominations of settlements or parts thereof in different languages. Hereafter, we conduct a detailed examination of the relationship between the place of birth and current residence of foreign citizens coming to Hungary from neighbouring countries at Nomenclature of Territorial Units for Statistics (NUTS3) level2, using the 2001 and 2011 census data (Tóth 2013). As for Ukraine, instead of 1 As a first step, the two databases have been connected by all the five variables. With this method, most records could be corresponded to each other. In the second step, the non-connected rows have been unified with the help of four-element keys (citizenship, the settlement of the Hungarian residence, and two-element subsets of all three- element variable sets). In some remaining (not connecting) cases, we have used the less differentiated level of district instead of the settlement of the Hungarian residence, thus elaborating five and four-element keys with this method. The application of this method resulted in formation of the new, interconnecting database. 2 This territorial classification is available for most countries, with the exception of Serbia and Ukraine. In case of Serbia, the most probable NUTS3 division is applied (For the territorial division of the Serbian Republic, the classification published in the 2011 Statistical Yearbook has been used. For detailed information, see: http://www.media.srbija.gov.rs/medsrp/dokumenti/SGS2011_cyr.pdf), while in Ukraine such a classification does not exist. The ‘oblaszt’ level is more integrated and the ‘rayon’ is more detailed than this classification. A more defined division will be applied to Transcarpathia as it has a distinguished role within Ukraine (the vast majority of migrants come from here). Regional Statistics, Vol 6, No 2. 2016: 95–113; DOI: 10.15196/RS06205 Migration settlement networks in the Carpathian Basin, 2001–2011 97 the whole country – due to its large size – only Transcarpathia has been involved in the study since nearly 90% of Ukrainian migrants arrive from this region. From the 161 regions established, in this manner significant concentrations can be observed in the matrix of migrations to the 19 Hungarian counties and Budapest. Extracting those region pairs, which contribute to the total turnover and have more than 0.5% of the total migration, we get a much tighter group than the previous one. Thus, 43% of migrations are concentrated in 1% of all the matrix cells (42x20=840) in 2011. This proportion increased by nine percentage points until 2011. Central Hungary was the most attractive destination for those arriving from the counties of Transylvania in 2011. Nearly, 4% of all immigrations from Romania to Hungary took place between Harghita county and Budapest, while the share of moves between Mures County and Budapest was 3%. The border areas were of considerable importance as well, which can be explained partly by the phenomenon of circulation migration (Fercsik R 2008, Elijah S. et al, 2009, Illés S. et al, 2009) and partly by the ease that borders offer to migrants for maintaining a connect with family members staying at home (Rédei M 2007). Intensive flows (Anderson et al, 1999, Baranyi, B. et al. 2004, N. Hansen 1977; Van Geenhuizen, M. et al, 2001) and transnational areas were formed (Melegh, 2011) between the bordering counties. Among them, the most significant movements included North Backa, North Banat-Csongrád (2,37% and 1.16%) as well as Beregovo and Uzsgorod rayon-Szabolcs-Szatmár-Bereg county (2.42% and 1.35%). During the same period, a population of mostly Hungarian ethnicity came from the large emitting areas to Hungary. The number of pairs of regions, which are affected by more than 0,5% by the migrations from neighbouring countries to Hungary showed an increase by 2011. The spaces of migratory relations increased in Hungary, with remoter parts of the neighbouring countries joining the source areas, thereby increasing the regional role of Trnava, Bratislava, Košice, Nitra districts, and Vienna’. (Estélyi K. et al. 2006). The significance of Budapest and Pest County further strengthened, while the weight of migration of Szabolcs-Szatmár-Bereg county and, to a certain extent, of Csongrád weakened. Generally, in these counties the number of migrants also increased, but not as much as in Central Hungary. By 2011, the migration ratios between Harghita, Mures, and Budapest continued to increase (3.8% and 4.5%, respectively). Generally, the cross-border relationship levels retreat. Often, those arriving from the proximity of the borders do not migrate to the neighbouring region of Hungary but to Central Hungary, which provides better structural (labour market and income) facilities. Regional Statistics, Vol 6, No 2. 2016: 95–113; DOI: 10.15196/RS06205 98 Áron Kincses – Lajos Bálint Table 1 Proportions of major migration flows*, 2001 (%) Sza- Foreign/ Buda- Bács- Csong- Hajdú- bolcs- Hungarian Békés Fejér Pest pest Kiskun rád Bihar Szatmár- counties Bereg Bihor 1.63 0.18 0.39 0.10 0.18 0.70 2.16 0.20 Brasov 1.25 0.08 0.05 0.11 0.09 0.03 0.21 0.03 Cluj 2.55 0.14 0.19 0.18 0.50 0.35 0.74 0.20 Covasna 1.32 0.12 0.18 0.07 0.16 0.33 0.46 0.08 Harghita 3.75 0.46 0.51 0.65 0.41 0.40 2.61 0.30 Maramures 0.92 0.06 0.05 0.06 0.12 0.05 0.43 0.09 Mures 2.77 0.44 0.47 0.65 0.29 0.32 0.91 0.22 Salaj 1.00 0.06 0.04 0.04 0.16 0.23 0.25 0.06 Satu Mare 1.92 0.11 0.15 0.16 0.34 0.72 0.37 0.84 Timis 0.07 0.020.06 0.11 0.03 0.03 0.67 0.02 Trnava district 0.51 0.00 0.02 0.02 0.00 0.02 0.12 0.03 Berehove rayon 0.85 0.08 0.05 0.06 0.13 0.34 0.28 2.42 Vynohradiv rayon 0.56 0.10 0.06 0.06 0.01 0.15 0.21 0.90 Uzshorod rayon 0.67 0.06 0.11 0.03 0.10 0.30 0.42 1.35 Juznobacki district 0.79 0.36 0.10 0.85 0.08 0.02 0.08 0.02 Severnobacki district 0.71 0.50 0.30 2.37 0.11 0.05 0.12 0.01 Severnobanatski district 0.68 0.48 0.14 1.16 0.45 0.13 0.32 0.01 * Total number of those coming from the regions of all neighbouring countries to all Hungarian counties is 100%.